2023 Volume 64 Issue 2 Pages 398-405
High strength AA6XXX series aluminium alloys have become the material of choice for body structures of lightweight automotive vehicles. The properties can be tailored through thermomechanical processing and strict adherence to impurity level. Although some nickel-containing AA6XXX Al alloys were introduced previously, the effect of Ni addition is still out of significant attention. Presently, an AlMgSiCu alloys with 0.1, 0.2 and 0.3 wt.% Ni addition was studied along with thermodynamic analysis. The predicted intermetallic phases in the as-cast base alloy were Al15(Fe, Mn)3Si2, Al5FeSi, Mg2Si and Q phases and also Al3Ni and Al3(CuNi)2 in the as-cast Ni-added alloys. The Al3(CuNi)2 phase with irregular shape was found in all as-cast Ni-added alloys and Al9FeNi phase was also found around dendritic cells in the as-cast alloy containing 0.3 wt.% Ni. After homogenization the Al3(CuNi)2 phase became much leaner in Cu content showing a possible phase transformation occurred. It was found that even at the low Ni content the excess Ni atoms could interact with neighbouring Fe atoms upon heat treatment to form Al9FeNi phase and the Cu atoms were released into the matrix. Thus, the Ni addition could not only lead to the loss of Cu atoms available for subsequent age hardening but also reduce Si entrapment associated with Fe-rich intermetallics. Moreover, improved circularity of intermetallics was found due to appearance of the Al9FeNi phase that could potentially mitigate the detrimental effects of Fe.
High strength Al–Mg–Si–Cu series aluminium alloys have become the key lightweight material choice for automotive body structures due to their excellent processability and hardening response. Their main advantage is achieved by the effect of Cu which is either enriched in the β′′ precipitate1) or produces additional Cu-rich GP zones and L, S, C, QP, QC and Q′ precipitates, as needle or plate-like precursors to quaternary Q phase.2) This makes the Al–Mg–Si–Cu alloys much stronger than their Cu-free counterparts.
It is well-known that the properties of wrought aluminium alloys are controlled by the thermomechanical processing and the impurity tolerance that is currently under evaluation towards more recyclable development.3) Some studies confirmed that the AA6xxx Al alloys could lose some strengthening potentials due the entrapment of the main alloying elements (e.g. Mg and Si) by the impurities.3–6) For instance, iron, the most common recycling-origin contaminant inherent from recycling process is always present as insoluble intermetallics such as β-Al5FeSi, α-Al8Fe2Si, α-Al15(Fe, Mn)3Si2 or π-Al8FeMg3Si6 that may reduce Si supersaturation level.3,5) Also, Mg atoms can be consumed, for example, by the formation of binary phases together with low-melting elements like Pb or Bi.6) Loss of Cu has been reported to a much less extent but it is most likely to be caused by Ni impurity. While some studies proposed new Ni-containing Al alloys,7–9) very little information is available about Ni effects in Cu-containing AA6xxx alloys.
Though a lion’s share of existing Cu-rich phases dissolves during homogenisation, excess transition elements such as Fe and Ni could bring the formation of insoluble phases. In terms of Fe, the Al7CuFe2 phase normally occurs in 2xxx or 7xxx Al alloys but such phases have never been reported in the Al–Mg–Si–Cu alloys.10) On the other hand, the effect of Ni impurity still requires attention. This element can either significantly reduce Cu solubility due to the formation of Al3CuNi or Al7Cu4Ni phases10–12) or improve ductility due to the formation of compact Al9FeNi phase instead of needle-like β-Al5FeSi phase.7,8,10) Ni impurity can be picked up during both production of primary metal or during remelting of post-consumer scrap. The latter can be possible if scrap pieces include wrought alloys, for instance, AA2618, AA4032 or AA8001, which might contain about 1 wt.%Ni.10) This raised some studies on the effect of minute Ni impurity in commodity alloys such as AA6060 and AA6063.13,14) It has been reported that when the Ni content is increased to 500 ppm it tends to either partition in the α-Al15(Fe, Mn)3Si2 phase or produce the Al9FeNi phase which can dissolve up to 4 wt.% Si. It was found that in both cases Ni addition triggers the formation of intermetallics that are less electronegative than Al-, hence reducing resistance to cosmetic corrosion. It was found that adding 0.03 wt.%Ni to Al–Mg–Si–Cu–Fe–Zn–Ni alloy could enhance the β-Al5FeSi → α-Al15(Fe, Mn)3Si2 transformation presumably due to the presence of Al9FeNi phase7) which, however, was not detected. And for the as-cast Al–1.2Mg–0.7Si–1.0Cu–0.9Fe–0.9Ni–0.2Sc–0.2Zr alloys, the Cu element was found to be segregated as the Al2Cu phase along the dendritic cells boundaries and Ni is fully incorporated in Al9FeNi phase,15) that could become spherical shape after homogenisation. According to Refs. 14), 16) the Al9FeNi phase might dissolve up to 2.5 wt.%Cu and 4 wt.%Si. Apparently, it can make the phase composition more complicated and affect the strengthening potentials. As a result, the current knowledge about the effect of Ni impurity on high strength AA6xxx alloys remains ambiguous. Preliminary trials are presently taken to estimate the effect of Ni on high strength Al–Mg–Si–Cu alloy along with discussions about its influence on phase composition, microstructure, solute level and consequent hardening potential.
To obtain an idea about the effect of Ni on the phase composition of high strength Al–Mg–Si–Cu alloys, Thermo-Calc software and the TCAl4.0 database were employed.17) The CALPHAD approach implemented in the software has been established to give reliable results to predict the microstructural behaviour even though it primarily considers equilibrium conditions assuming perfect mixing in liquid and solid phases. The reference material was a high-strength AlMgSiCu alloy with a composition close to the AA6111 Al alloy.18) To understand Ni effect, the alloys containing 0.1 wt.%Ni, 0.2 wt.% and 0.3 wt.%Ni (hereinafter referred to as 0.1Ni, 0.2Ni and 0.3Ni alloys) were produced along with the reference alloy (Table 1). The melting was carried out in an electric furnace at 720°C and TP-1 castings were obtained according to the standard procedure.19) Before pouring, the commercial grain refiner Al–5Ti–1B was added at the level of 1 wt.% of the total charge. The samples were taken at a height of 38 mm from the bottom of TP-1 samples which gives a cooling rate replicating that in the half-radius of 180 mm DC cast billet. According to Ref. 19), the cooling rate of the AA6063 Al specimen taken at the same position was 3.9 K/s. The microstructures were analysed in as-cast and T4 conditions (homogenisation annealing followed by solution treatment by quenching in water and natural ageing for 7 days) conditions using a Zeiss Leo 1455VP scanning electron microscope (SEM) equipped with an energy dispersive x-ray sensor (EDS). Quantitative metallography including volume fraction, size and circularity (4π·area/perimeter2) was carried out using ImageJ software.20) At least 1000 particles per condition were collected for statistically representative results. To examine the precipitation hardening response, the samples were artificially aged at 170°C with different aging time. Eddy-current electrical conductivity (EC) and Vickers microhardness (HV withload of 1 kg and dwell of 15 s) tests were employed to check the strengthening after each processing step.
According to Ref. 21), a multi-component system Al–Mg–Si–Cu–Fe–Mn should be considered to investigate the solidification behaviour of the present alloy. In total eight reactions can take place during solidification and five phases in the as-cast microstructure including β-Al9Fe2Si2 (hereinafter β-Fe), α-Al(Fe,Mn)Si (hereinafter α-Fe), Mg2Si, Q-AlMgSiCu and eutectic (Si) are formed. The last three phases appear because of incomplete diffusion or non-equilibrium solidification at the last stage of solidification at around 500°C lower than the equilibrium solidus (∼570°C) and depends mainly on Si content. Hence, an appropriate homogenisation treatment would be required to dissolving non-equilibrium eutectic microstructures and break interconnectivity of Fe-rich intermetallics. This study focuses on the investigation of whether the notoriously thermally-stable Ni-rich intermetallics can change their shapes and compositions during thermal treatment.
The polythermal section (Fig. 1) shows the equilibrium phase composition of the experimental alloys between 350 and 700°C. It can be seen that in all studied compositions, the liquidus temperature (∼650°C) is nearly constant as well as the nucleation temperatures of the α-Fe (∼630°C), Mg2Si (∼517°C) and β-Fe (∼425°C) phases. Meanwhile, the solidus temperature reduces due to the formation of the Ni-rich intermetallic phase. Specifically, in the reference alloy, the last liquid disappears at 580°C followed by the establishment of the α-Fe phase. This temperature sharply decreases to 567°C when Ni content reaches 0.09 wt.%, the maximum solid solubility in Al matrix. Between 0.09 wt.% Ni and 0.175 wt.% Ni, the Al3(CuNi)2 phase appears, and the solidus temperature slightly increases to 570°C. Further Ni addition does not change the solidus but brings the Al3Ni phase into the equilibrium eutectic mixture. The Al3Ni phase nucleates after the α-Fe phase and its nucleation temperature is increasing with Ni, which is accompanied by a slight Fe depletion in the α-Fe phase because the Al3Ni phase can dissolve some Fe atoms. However, there is no sign of Al9FeNi phase that is either due to database incompleteness or lack of Fe to preferentially interact with Ni to form the ternary aluminide. For example, the Al9FeNi phase is commonly observed in the heat-resistant AA2618 Al alloy with up to 2.7 wt.%Cu and also an appreciable amount of Ni and Fe contents (up to 1 wt.% of each).10)
Polythermal section of Al–Mg–Si–Cu–Fe–Mn–Ni system at 0.7%Mg, 0.8%Si, 0.8%Cu, 0.2%Fe, 0.5%Mn and varying Ni and temperature (Al15 - α-Fe, Al9 - β-Al9Fe2Si2).
The Cu content in Al matrix and the fraction of intermetallic phases at solution treatment temperature in the Al–Mg–Si–Cu alloy with different Cu contents and Ni impurity levels are plotted in Fig. 2. The α-Fe fraction (∼1.56 vol.%) is nearly stable in all studied alloys while that of the Ni-rich phases is increasing with Ni content. Interestingly, in the alloys containing up to 0.7 wt.%Cu there is no consumption of Cu by the Ni-rich phases because Ni is only bonded to the binary Al3Ni phase that can approach 0.38 vol.% at 0.3 wt.%Ni. Further increasing the Cu content promotes the formation of the Al3(CuNi)2 phase at the expense of the Al3Ni phase and Cu solute. At 0.1 wt.%Ni the decrease in Cu solute of 0.03 wt.% might be negligible whereas for 0.2 and 0.3 wt.% Ni it could consume up to 0.07 wt.% and up to 0.15 wt.% Cu solute, respectively. This Cu loss or depletion in Al matrix cannot be compensated by additional Cu within the specification (e.g., max. 0.9 wt.% Cu in AA6111 Al alloy) because excess Cu solute may interact with Ni until the Al3Ni phase completely disappears and Al3(CuNi)2 fraction is established.
Equilibrium Cu content in Al matrix and the equilibrium fraction of intermetallic phases at solution treatment temperature in the Al–Mg–Si–Cu alloy with different Cu contents and Ni impurity levels.
In general, the above discussion signifies that the Ni impurity at a given Fe content (∼0.2 wt.%) in the high-strength Al–Mg–Si–Cu alloy can lead to sizeable growth of intermetallics volume fraction and a decrease in Cu solute level in the Al matrix. Table 1 summarises the said parameters at the solid solution temperature. According to the equilibrium calculations, the alloys with different Ni contents can have completely different microstructures due to the formation of ternary eutectic L → (Al) + α-Fe + Al3(CuNi)2 (567°C) or quaternary eutectic L → (Al) + α-Fe + Al3(CuNi)2 + Al3Ni (570°C). The latter can contain either excess of Al3(CuNi)2 over the Al3Ni phase in the alloys with 0.1 wt.%Ni and 0.2 wt.%Ni or excess of Al3Ni over Al3(CuNi)2 in the 0.3 wt.%Ni alloy. One should note that potential partition of Ni and Cu in the Fe-rich phases after real (non-equilibrium) solidification might bring variance in phase composition. For instance, according to Scheil-Gulliver simulation, which considers partially non-equilibrium condition, gives non-equilibrium solidus rising with Ni from 510°C in the reference alloy to 515°C in the 0.3Ni alloy. This phenomenon can be explained by change in multicomponent eutectic reaction from L → (Al) + α-Fe + Al2Cu + (Si) + Q to L → (Al) + α-Fe + Al7Cu4Ni + (Si) + Q, i.e. there is an formation of Al7Cu4Ni instead of Al2Cu phase after addition of nickel.
3.2 As-cast microstructuresAs displayed in Fig. 3, all alloys are composed of grey (Al) dendritic cells surrounded by a number of various phases. The Mg2Si and Q-AlMgSiCu phases represent the main constituents of high-strength Al–Mg–Si–Cu alloys and are visible but the change in their fraction is insignificant. They should be dissolved during solid solution treatment for high supersaturation level to form nanoscale strengthening precipitates. In addition, the α-Fe phase with elongated, blocky, or Chinese-script shapes is dominating among the second phase particles in the as-cast alloys.
Microstructures of the as-cast alloys: (a) – Base alloy; (b) – 0.1Ni alloy; (c) – 0.2Ni alloy; (d) – 0.3Ni alloy.
In the 0.1Ni alloy, the bright Al3(CuNi)2 intermetallics have a form of short needles mainly associated with the grey α-Fe phase. In the 0.2Ni alloy, the fraction of Al3(CuNi)2 visually increases and some Ni elements is found associated with the Q-AlMgSiCu phase. The latter phenomena might be explained by a coupled formation of the Al7Cu4Ni phase and Q-AlMgSiCu phase according to the eutectic reaction L → (Al) + α-Fe + Al7Cu4Ni + (Si) + Q. Similar reaction L → (Al) + (Si) + Al7Cu4Ni + Al2Cu + Q at about 505°C was observed in Ref. 22). Surprisingly, the 0.3Ni alloy contains Fe- and Ni-rich phase, which is of the same colour as the α-Fe but has a more curved shape that might account for its better spheroidisation response during heat treatment. According to Refs. 10), 16), the Al9FeNi phase has wide Fe (4.5 wt.% to 14 wt.%) and Ni (28 wt.% to 18 wt.%) contents and Al3Ni phase can dissolve no more than 1 wt.%Fe. Apparently, the newly-appeared phase is much closer to Al9FeNi according to EDS results (Table 2). Preliminary assessment could assume that such phase is likely appeared due to excess of Ni over Fe in the 0.3Ni alloy, which reacted with free Fe during solidification at the expense of β-Fe that forms at a much lower temperature. Thus, the experimental results do not necessarily agree with the thermodynamic data.
Results of EDS analysis of representative intermetallic particles containing Ni and Fe are presented in Table 2. Since the incident electron beam of EDS also penetrated the Al matrix to a certain extent, it can bring some scatter in the results but still allows for the distinction of different phases. The α-Fe phase, the main constituent in the microstructure, contains 1.6–2 wt.%Cu in all alloys, which is in agreement with previously published data23) that indicate that Cu can replace Si sites in the Al15(FeMn)3(Si, Cu)2 compound. It is interesting to note that with an increase in Ni, there is more Ni in the area corresponding to the α-Fe phase, e.g. 1.2 wt.%Ni in the 0.1Ni alloy and 1.93 wt.%Ni in the 0.3Ni alloy. This might be attributed to the replacement of Fe atoms in the α-Fe phase by Ni and can account for the alteration of Fe composition. The latter is especially prominent in the 0.3Ni alloy due to the formation of Al9FeNi phase. Potentially, Ni dissolution in the Al15(FeMn)3(Si, Cu)2 phase can further reduce corrosion resistance because of alteration of dissolution potential.13) The same could be expected from the cathodic Al3(CuNi)2 phase the formation of which brought a sizeable increase in volume fraction of intermetallics. As for the shape and size of intermetallics, there is a decrease in the circularity and an increase in the size and number of large intermetallics. This trend accounts for the frequently elongated shape of Ni-rich phases (especially Al3(CuNi)2) and their interconnectivity with the α-Fe phase. From this viewpoint, they will deteriorate the mechanical performance of the alloy and a suitable homogenisation practice should be applied to improve their shape and size parameters.
3.3 Microstructures after heat treatmentIn the metallurgy of AA6xxx alloys, homogenisation treatment can help significantly improve the processability and properties of wrought products due to the reduction in residual stresses from casting and promotion of diffusion-based tuning of microstructures.24,25) It can be seen from Fig. 4 that all experimental alloys encountered concurrent transformations intrinsic for high strength Al–Mg–Si–Cu alloys25) despite increased Ni impurity. Firstly, there is no undissolved Q and Mg2Si phase, hence the proper homogenisation temperature is chosen to provide high Mg, Si and Cu supersaturation in (Al) that is also supported by the polythermal section (Fig. 1(a)). The dissolution of the said phases may cause a slight decrease in the volume fraction of intermetallics and an increase in the average size of the remaining constituents. Secondly, barely perceptible but still traceable submicron-scale Mn-containing dispersoids are precipitated in the centre of dendritic cells to serve as grain-pinning agents during the thermomechanical treatments. Finally, high-temperature treatment promotes the spheroidization of insoluble intermetallics and possible change in their chemical composition.
Microstructures of the experimental alloys after heat treatment: (a) – Base alloy; (b) – 0.1Ni alloy; (c) – 0.2Ni alloy; (d) – 0.3Ni alloy.
Microstructures of the reference alloy (Fig. 4(a)) and the 0.1Ni alloy (Fig. 4(b)) show similar shapes of the α-Fe phase and spheroidization response as it can be seen from the circularity values. According to calculations, at the chosen temperature only around 0.05 wt.%Ni can be dissolved in solid solution, which means that some residual Ni in the experimental alloys can still be remaining in the form of intermetallics. In the 0.1Ni alloy, the amount of these intermetallics decreases significantly thus agreeing well with the calculated data (Table 1). Nevertheless, some minor amount of residual needle-like Ni-rich intermetallics is still present in the 0.1Ni alloy contributing to the increase of the volume fraction and average size. However, the Ni-rich intermetallics become significantly leaner of Cu solute upon heat treatment in all experimental alloys, which could be observed despite the semi-quantitative nature of the EDS analysis (Table 3). One should note that similar Cu-lean phase can be observed even in the wrought product made from similar alloy with Ni content >1 wt.% which was studied in Ref. 26). Possible dissolution of the Al3(CuNi)2 phase during thermal exposure and change in chemical composition was previously reported.27) According to Ref. 11), the Al7Cu4Ni phase can be transformed into Al3Ni and Cu solutes in Al matrix. In Ref. 28) both the Al3(CuNi)2 and Al7Cu4Ni phases were reported as thermodynamically unstable phases at high temperatures, and could be transformed into Al9FeNi and/or Al3Ni phases upon heat treatment. One should note that these findings were observed in the Al–Si and Al–Cu–Si casting alloys and never reported in AA6xxx Al alloys. Apparently, the decomposition of the Al3(CuNi)2 phase can bring a less detrimental effect of Ni impurity as it appears not to significantly consume Cu solute. Hence, this effect is worth further studying. In this study, it is assumed that the new phase is close to the Al3Ni phase as it shows a substantial increase in Ni content with high Ni impurity level in the alloy. In the 0.3Ni alloy, the Al3Ni phase is coarser than that in the 0.1Ni and 0.2Ni alloys, which agrees with their increased volume fraction as calculated. Furthermore, in the 0.2Ni alloy (Fig. 4(c)) some Ni is found to be either incorporated or located nearby the α-Fe phase. It can be assumed that the Al9FeNi phase appears due to free Ni interacting with Fe atoms in the neighbouring intermetallics. A similar phenomenon was observed in Ref. 8) where during heat treatment the β-Fe phase converted to the Al9FeNi phase at the β-matrix interface. Hence, an appreciable spheroidization of the notoriously-known needle-like β-Fe phase could be achieved resulting in better mechanical properties. In this study, the 0.2Ni and 0.3Ni alloys have much higher circularity but also higher volume fraction and size of intermetallics that can be explained by spheroidization and coarsening of the branches of the α-Fe + Al9FeNi mixture.
The composition of the Al matrix (Table 4) has been measured on at least 20 points in the intermetallics-free areas. It can be seen that the solubility of Mg and Si is relatively stable in all the experimental alloys, which can be explained by no trapping of these elements by Ni impurity. Meanwhile, the Cu level is less stable showing a decreasing trend with an increase in Ni content. According to the microstructures of the solution-treated alloys, the Cu solute is incorporated in all intermetallic phases α-Fe, Al3Ni and Al9FeNi to a certain extent. Thus, it can be assumed that the Cu level should be decreased with an increase in the fraction of the said phases. Nevertheless, the α-Fe and Al3Ni phases can dissolve around 2 wt.%Cu whereas the Al9FeNi phase only 1 wt.% which agrees with the literature data.22) Thus, we can see a significant decrease in Cu solubility in the 0.1Ni alloy which includes only α-Fe and Al3Ni phases, a slight recovery in the 0.2Ni alloy which includes some Al9FeNi phase, and a significant drop in the 0.3Ni alloy with an inadequate fraction of Al3Ni phase. That difference in Al3Ni fraction between 0.2Ni and 0.3Ni alloys was shown by thermodynamic calculations as well (0.211 vol.% vs. 0.047 vol.%).
Changes in the Vickers hardness (HV) and electrical conductivity (EC) values after solid solution and ageing treatments represents the extent of decomposition of the solid solution followed by the formation of hardening precipitates. In high-strength AA6xxx Al alloys, it takes into account the formation of Mn-rich dispersoids during pre-solution high-temperature exposure and the formation of GP zones during natural ageing. According to Fig. 5(a), in the as-cast condition, an increase in Ni content can reduce the hardness from ∼70 HV to ∼63 HV which can be explained by more Cu entrapment in the Al3(CuNi)2 phase. This is supported by EC values which are reduced only slightly because they are influenced primarily by Mn supersaturation as compared to Mg, Si and Cu solutes.16) As a result, a substantial increase in EC can be seen in the T4 condition due to the decomposition of (Al) solid solution and the formation of Mn-rich dispersoids, whereas the hardness is mainly stabilised at the level of ∼65 HV for all alloys. Maximum HV values can be achieved after the exposure at 175°C for 20 h, whereas the EC value is increasing gradually most likely due to precipitate coarsening. It can be seen that only 0.1Ni alloy shows inferior HV as compared to other alloys which represent a similar trend. It can be assumed that the 0.1Ni alloy shows a decrease in HV due to the loss of Cu solute associated with both the α-Fe phase and Al3Ni phase, whereas in the 0.2Ni and 0.3Ni alloys the formation of the Al9FeNi phase could reduce the fraction of Al3Ni and α-Fe phase followed by both reduction of Cu entrapment and increasing free Si. In this case, despite (Al) solid solution is leaner with Cu in the 0.3Ni alloy, and its strength might be compensated by free Si as a powerful strength contributor.28) Thus, for the prospects, it is worth verifying whether Ni impurity can reduce Si entrapment associated with increased Fe impurity in AA6xxx Al alloys produced out of scrap. In the EC curve, all the alloys show similar behaviour except the 0.3Ni alloy with just slightly higher values (24.9 vs. 24.5 MS/m after 175°C, 20 h).
Evolution of the microhardness (a) and electrical conductivity (b) during heat treatment of the experimental alloys.
In this work, the AA6xxx Al alloys with different Ni impurity levels are investigated in the as-cast, T4 and T6 conditions using thermodynamic modelling and experimental validation. The following conclusions can be drawn:
The authors gratefully acknowledge the grant to Brunel University received from the Advanced Propulsion Centre through Innovate UK that supported the work programme presented.
